Workshop: Exploring the InsideAirBnB dataset - EDA

Preamble

# Clear workspace
rm(list=ls()); graphics.off() 

### Load packages
library(tidyverse) # Collection of all the good stuff like dplyr, ggplot2 ect.
library(magrittr) # For extra-piping operators (eg. %<>%)
library(skimr) # For nice data summaries

The InsideAirBnB data

Instroduction

• The data is sourced from the Inside Airbnb which hosts publicly available data from the Airbnb site.
• Interactive visualizations are provided here

The dataset comprises of three main tables:

• listings - Detailed listings data showing 96 atttributes for each of the listings. Some of the attributes which are intuitivly interesting are: price (continuous), longitude (continuous), latitude (continuous), listing_type (categorical), is_superhost (categorical), neighbourhood (categorical), ratings (continuous) among others.
• reviews - Detailed reviews given by the guests with 6 attributes. Key attributes include date (datetime), listing_id (discrete), reviewer_id (discrete) and comment (textual).
• calendar - Provides details about booking for the next year by listing. Four attributes in total including listing_id (discrete), date (datetime), available (categorical) and price (continuous).

Load data

listings <- read_csv('http://data.insideairbnb.com/denmark/hovedstaden/copenhagen/2020-06-26/data/listings.csv.gz')
listings %>% head()

calendar <- read_csv('http://data.insideairbnb.com/denmark/hovedstaden/copenhagen/2020-06-26/data/calendar.csv.gz')
calendar %>% head()

reviews <- read_csv('http://data.insideairbnb.com/denmark/hovedstaden/copenhagen/2020-06-26/data/reviews.csv.gz')
reviews %>% head()

# The geodat of the hoods comes as a geojson, so we need the right package to load it
library(geojsonio)
neighbourhoods_geojson <- geojson_read( 'http://data.insideairbnb.com/denmark/hovedstaden/copenhagen/2020-06-26/visualisations/neighbourhoods.geojson',  what = "sp")

Some preprocessing

listings %<>% mutate(price = price %>% parse_number()) 

Ad-hoc investigation

Problem 1: Professional hosts & their characteristics

listings %>%
  count(host_id, sort = TRUE)

Where are they?

listings %>%
  filter(host_id == 187610263) %>%
  count(neighbourhood_cleansed, sort = TRUE)

Dummy for professional host

listings %<>%
  group_by(host_id) %>%
  mutate(host_professional = n() >= 5) %>%
  ungroup()

listings %>%
  group_by(host_professional) %>%
  summarise(review = review_scores_rating %>% mean(na.rm = TRUE),
            price = price %>% mean(na.rm = TRUE))

-> Profressional hosts charge more…

listings %>%
  group_by(neighbourhood_cleansed, host_professional) %>%
  summarise(review = review_scores_rating %>% mean(na.rm = TRUE)) %>%
  pivot_wider(names_from = host_professional, values_from = review)

-> This is true everywhere, yet in some hoods mnore tghan in others

Problem 2: Length of description -> Satisfaction

listings %<>%
  mutate(desc_lenght = description %>% str_count('\\w+')) %>%
  mutate(desc_long =  percent_rank(desc_lenght) > 0.9 )

listings %>%
  group_by(desc_long) %>%
  summarise(review = review_scores_rating %>% mean(na.rm =TRUE))

-> No overall effect

P3: Best party place

listings %<>% 
  mutate(party_place = accommodates >= 10) 

listings %>% 
  filter(party_place == TRUE) %>%
  group_by(neighbourhood_cleansed) %>%
  summarize(n = n(),
         review = review_scores_rating %>% mean(na.rm = TRUE),
         price = price %>% mean(na.rm = TRUE),
         price_pp = (price / accommodates) %>% mean(na.rm = TRUE)) %>%
  arrange(desc(n))

If you are on a tight budget, best go to Amager-Vest.

Geoplotting

Interactive map

library(leaflet)

listings %>% leaflet() %>%
  addTiles() %>%
  addMarkers(~longitude, ~latitude,
             labelOptions = labelOptions(noHide = F),
             clusterOptions = markerClusterOptions(),
             popup = paste0("<b> Name: </b>", listings$name, 
                            "<br/><b> Host Name: </b>", listings$host_name, 
                            "<br> <b> Price: </b>", listings$price, 
                            "<br/><b> Room Type: </b>", listings$room_type, 
                            "<br/><b> Property Type: </b>", listings$property_type
                 )) %>% 
#  setView(-74.00, 40.71, zoom = 12) %>%
  addProviderTiles("CartoDB.Positron")

Choropleth Plots

# Using broom to tidy the geojson
library(broom)
neighbourhoods_tidy <-  neighbourhoods_geojson %>%
  tidy(region = "neighbourhood")

neighbourhoods_tidy %>% glimpse()

Rows: 6,658
Columns: 7
$ long  <dbl> 12.63094, 12.63126, 12.63221, 12.63160, 12.63154, 12.63153, 12.63153, 12.63153, 12.63157, 12.63158, 12.6…
$ lat   <dbl> 55.67050, 55.67028, 55.66961, 55.66943, 55.66941, 55.66940, 55.66939, 55.66930, 55.66926, 55.66924, 55.6…
$ order <int> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 2…
$ hole  <lgl> FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE,…
$ piece <fct> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,…
$ group <fct> Amager st.1, Amager st.1, Amager st.1, Amager st.1, Amager st.1, Amager st.1, Amager st.1, Amager st.1, …
$ id    <chr> "Amager st", "Amager st", "Amager st", "Amager st", "Amager st", "Amager st", "Amager st", "Amager st", …

neighbourhoods_tidy %>%
  ggplot(aes(x = long, y = lat, group = group)) +
  geom_polygon() +
  theme_void() +
  coord_map()

neighborhood_agg <- listings %>%
  group_by(neighbourhood_cleansed) %>%
  summarise(n = n(),
            price_mean = price %>% mean(na.rm = TRUE),
            review_mean = review_scores_rating %>% mean(na.rm = TRUE))
  

neighbourhoods_tidy %<>%
  left_join(neighborhood_agg, by = c('id' = 'neighbourhood_cleansed'))

Number of places

neighbourhoods_tidy %>%
  ggplot(aes(x = long, y = lat, group = group, fill = n)) +
  geom_polygon() +
  theme_void() +
  coord_map()

Prices

neighbourhoods_tidy %>%
  ggplot(aes(x = long, y = lat, group = group, fill = price_mean)) +
  geom_polygon() +
  theme_void() +
  coord_map()

Review scores

neighbourhoods_tidy %>%
  ggplot(aes(x = long, y = lat, group = group, fill = review_mean)) +
  geom_polygon() +
  theme_void() +
  coord_map()